Practical use of branching in Mercurial?

Question

CVS

Gentlemen, in a content management system like Mercurial, there are many ways to "branch" history: branch (nameless and named), tag , bookmarks , clone .
Share your experience on how real projects of various sizes use these features for:
1. Maintaining major versions (v1, v2, v3)
2. Minor versions (v1.1, v1.2, ...)
3. Sub-versions (v1.1a , v1.1b, ...)
4. Stable (stable) and developed (dev) branches
5. Bug fixes (both minor and major)
6. Adding feature bugs (also of varying degrees of severity)

Answer 1

Look at the article nvie.com/posts/a-successful-git-branching-model/e
on the example of GIT - but it does not change the essence. In my opinion, the fewer branches, the better.
Here are a couple more articles
continuousdelivery.com/2011/05/make-large-scale-changes-incrementally-with-branch-by-abstraction/
martinfowler.com/bliki/FeatureBranch.html
We are developing one unstable branch - we create a new one before release , for example v3 and make it stable.
Other than that, there are no other branches (minor or sub-versions). For point 6, read the article FeatureBranch

Answer 2

Thanks for correcting. Unfortunately, the comment is no longer being edited.